In the synthesis of theories of practice and predictive processing (here and here), it becomes clear that what concepts like habitus and agency mean cannot be separated from what prediction and objective probability mean. Habitus formation is just another word for learning probabilities and for making predictions accordingly. This implies that the more exposure we have to certain objective probabilities, as in a field configuration, the more objective will become our predictions, the more we will do what is expected, and the better we will anticipate what those similarly ensconced in the field as us can recognize as, appropriately, a “next thing to do.” But as, in this sense, our action becomes almost entirely a form of social action, it also loses its purely subjective imprint, which in probabilistic terms, can be redefined as its randomness, its unpredictability, and surprise.
Rephrased this becomes the now classical question of whether a habitus can be broken or how, once acquired, habitus does not just become a determinative, “reproductive” mechanism. In probabilistic terms, this means that learned probabilities become limits of action and perception, which is not to say that action is fully predictable, only that we can find boundaries across which it will not go, perceptions which it will not perceive. For some, the fact that surprise is still possible negates the very presence of habitus (King 2000). For others, including Bourdieu (2000: 234) himself, the “relative autonomy of the symbolic order” is available to prevent the reproductive tendencies of habitus, by providing moments when objective probabilities themselves break and learned expectations no longer matter. This mimics objective situations (“critical moments”) when the loop really does break, and symbolic orders that have no specific grip on the world become just as likely as anything else to loop into the future. But we should not expect that this symbolic escape can be sustained as a replacement (Bourdieu 1988: chap. 5).
In this post, I want to suggest something different. In spaces of highly developed objective probability, “saturated with history” to such an extent that what can be done shows very little deviation from what has been done, an engagement with chance can allow for a form of habitus breakage. The space in question will be the field of art, which, so saturated with history, means that it is rife with reproduction and cliche. The interesting thing is that, in the examples below, some artists take note of this and engage chance in order to prevent themselves from being fully “objective” and therefore cliched. Such a standpoint on art-making can be described as follows, but this statement can easily be extrapolated toward a probabilistic theory of agency more generally:
If we consider a canvas before the painter begins working, all the places on it seem to be equivalent; they are all equally “probable.” And if they are not equivalent, it is because the canvas is a well- defined surface, with limits and a center. But even more so, it depends on what the painter wants to do, and what he has in his head: this or that place becomes privileged in relation to this or that project. The painter has a more or less precise idea of what he wants to do, and this pre-pictorial idea is enough to make the probabilities unequal. There is thus an entire order of equal and unequal probabilities on the canvas. And it is when the unequal probability becomes almost a certitude that I can begin to paint. But at that very moment, once I have begun, how do I proceed so that what I paint does not become a cliche? (Deleuze 2003: 93).
In the predictions that shape action, the artist will predict only what is already objective, which is to say, what is historical and has already been done. They will do this even as they try to do what is “improbable” as this is associated with artistic “worth” as the performance of creativity. As this suggests, to recognize an action as artistic requires some link to expectation: what one does is what is expected relative to those who are, likewise, oriented toward certain objective possibilities. More specifically, if we perceive only what we cannot predict, and if the experienced artist can predict most things about their own work, this means that they perceive less, which is a phenomenon that seems to follow from the acquisition of any form of expertise (Dreyfus 2004). Hence, art “experts” cannot help but fall into cliche and objective rationalization.
So how to break out of this? Consider the following description of what the art historian Yve-Alain Bois describes as “compositional art.” In this instance, the facts of composition are clear and speak directly to a specifically designed presence:
Composition is an intended, ordered relationship of discrete parts, a relationship that suggests-that at once builds and needs-an interiority, a solid, plotted depth that fills both the artist as intentional actor and the visual field, however flat, that underpins the painting: one is an analogue for the other … Composition names the pictorial relationship of discrete parts across a field, parts arranged according to a visual order that both underlies the whole, and of which it-the painting as a whole-is an individual instance, proof of laws and orders … The composed object, the structured or designed one, appears right and it appears necessary and specific-because the ordered relationship between the parts of the object structure a relationship between the object and the viewer, and more, between vision as conception and the world (Singerman 2003: 131).
Thus, what is composed is “flat,” “ordered,” “right” and “necessary.” It fulfills expectations. It can serve as proof of what art or a kind of art (a “style” or “genre”) should be. But all of this is exactly what non-composition attempts to work against, prevent and break. “Non-compositional art” is that which attempts to work against composition by finding its origin in what Bois calls the “motivation of the arbitrary.”
For Bois, in the most prominent examples of this kind of art-making, the technique used serves as a direct analogue to inviting chance in, because it is precisely the inclusion of non-compositional, chance mechanisms (invited in) that make it possible to find this very peculiar kind of motivation. An example of this is the approach taken by the artist Ellsworth Kelly and his effort, in the 1950s, to “escape the weight of Picasso.”
What is there left to do in the wake of someone who has invented everything? In this respect, at least, Kelly remains an American in Paris. Like Pollock and an entire generation along with him, he knows that, if he wants to accomplish anything, his first task is to escape the weight of Picasso. And since the latter cannot be outdone, the trick is not to try to outdo him. How? By eliminating the human figure, to start with which means not just refraining from producing effigies ex nihilo, but also not engendering them by displacement or condensation, since even at this little game, Picasso is unbeatable (he has a magician’s touch: not only does he have an infinite array of images in reserve, but he also has the ability to make something out of nothing) … But getting away from Picasso also entails renouncing all choice, all composition (Bois 1992: 16).
Thus, Kelly will set about evading Picasso by essentially trying to undermine his own compositional skills, or what we can appreciate as Kelly’s attempt to break his habitus as it can only reproduce objective possibilities that Picasso has done the most to define. Kelly first attempts a method of transfer, what he calls “transformed already-made” in which he simply sketches what is already there: “we are dealing here with a simple transfer—along the same lines as a photograph or an imprint … The raw material remains untransformed, but above all … it is almost untransformable: just try to make a human shape out of a grid!” (Bois 16). This includes sketching the lines of windows, seaweed, of objects seen in churches, of tennis courts, of street patterns, and most notably perhaps, of the awnings of a hotel (Awnings, Avenue Matignon). Thus in his answer to Picasso, Kelly can now say: “No need to play sorcerer’s apprentice; the figurative intention is not a necessary condition of esthetic transubstantiation; art can be made without transforming anything, without having to re-baptise; anything goes.” The goal, as Rosalind Krauss (1977) put it, is to produce an “index” without a “referent” as the epitome of the figurative opacity of the modernist movement.
Yet everything is still too compositional for Kelly, and so his habitus breakage engages with a second stage. This includes a more forthright practice with chance mechanisms. He begins to sketch under conditions of sensory deprivation (with his eyes closed, not looking at the paper) a clear effort to not allow himself to predict his own perceptions (Automatic Drawing, Pine Branches VI). Yet the problem he encounters are that the drawings come out as “too perfect, testifying to an exemplary motor coordination, or they are illegible failures that Kelly is not willing to accept as such, first because they look more like straightforward clumsiness than the product of chance, and also because, a priori, there is nothing to prevent one from detecting ‘unconscious’ images in them, and thus falling back—the old surrealist saw—into figuration” (Bois 25). What he realizes is that the “aleatory as such does not so easily give itself up to the eye … for the strictest order of chance to manifest itself, it has to be opposed to the strictest possible order …” And so Kelly makes a third attempt at habitus breakage, now more systematic. In a series of collages, Kelly would use chance mechanisms for their composition: he would cut a drawing into identically-shaped squares and then glue them in an order that tries to reproduce the original composition.
As Bois puts it, “it is impossible not to notice the unpredictable character of the joints … the geometrical distribution ends up being ever so slightly disrupted …” (25). In arguably his most famous efforts in this direction, Kelly starts with a modular grid and, then, pulling numbers from a box and darkening the corresponding blocks in the grid, produces a representation of light flickering on the surface of the Seine River. In another engagement with chance, Kelly uses a stock of gummed papers he happened to discover in a stationary shop, and taking the small squares of color, arranges them intuitively within a modular grid, “using no system or scientific method except to proceed progressively from the grid’s lateral sides toward the center.” In still other collages he would number the grids randomly, number the colors, and then put the colored squares into their corresponding boxes.
Thus, in these works, “the system presiding over [them] is in every case rigorous, but also conceived as a pure receptacle of chance. The notion of a systematic art flowed from the necessity of suppressing the arbitrariness of composition, but the arbitrariness presiding over the choice of system remained. Kelly’s extraordinary insight was to counterbalance that of the system with the still greater arbitrariness of chance, so as best to eliminate all subjective determination” (Bois (1999: 26). While this marks an attempt to get beyond and exterior to “subjectivity” in compositional decision-making, it is not an engagement with “end of author” approaches that would otherwise, in familiar fashion, emphasize structure or signification, over hermeneutics and depth. Rather, as Bois’ emphasizes, non-composition appears in technique, as practice, and in this capacity directly engages with chance as the antithesis to “subjective decision-making” and its compositional effects.
For other painters like Francis Bacon, the engagement with chance comes in the form of making “free marks” on the canvas, “so as to destroy the nascent figuration in it and to give the Figure a chance, which is the improbable itself. These marks are accidental, ‘by chance’; but clearly the same word, ‘chance,’ no longer designates probabilities, but now designates a type of choice or action without probability” (Deleuze 94). In the case of Three Studies for a Portrait of George Dyer (on Light Ground), the white mark drawn to shape Dyer’s face and strewn across it, is exactly this kind of “free mark” that seems to prevent the painting, based on a photograph, from being figurative. In Painting, Bacon describes how he first tried to draw a bird in a yard, but it turned out to be a man with an umbrella: “Well, one of the pictures I did in 1946, the one like a butcher’s shop, came to me as an accident. I was attempting to make a bird alighting on a field. … suddenly the lines that I’d drawn suggested something totally different, and out of this suggestion arose this picture. I had no intention to do this picture; I never thought of it in that way. It was like one continuous accident mounting on top of another. … I don’t think the bird suggested the umbrella; it suddenly suggested this whole image.” In this case, the marks that Bacon made turn out to be free marks, as they seem to elude the initial predictions he was making of his own work, such that he would eventually produce what he set out to. And so he, instead, fills the canvas with free marks.
What is notable about each of these “non-compositions” is that, as examples of inviting chance in, they also demonstrate the evident connection between probability and action. Merleau-Ponty (1964) describes the painter Henri Matisse, having been captured painting with a slow motion camera, amazed at how deliberate his brushstrokes appeared to be from this vantage, when from his vantage they were anything but deliberate. Using this engagement with time, it became clear that while viewing the recorded brush-strokes, the viewer (and Matisse himself) could witness an “infinite number of data … an infinite number of options” as possible. “Matisse’s hand did hesitate,” as Merleau-Ponty put its, choosing among “twenty conditions which were unformulated or even informulable for anyone but Matisse, since they were only defined and imposed by the intention of executing this painting which did not yet exist” (46).
It becomes evident that “art” or “painting” or “modernist painting” is not a structure of signification, though it still remains something that we can call objective. In this case, it is not Matisse’s creation; he, instead, works with it. But what he works with are probabilities, or historically-developed chances, ones that his hand transforms into something actual, that in some sense looks like it should, as he puts brush to canvas. This entire process, as Merleau-Ponty emphasizes, is not different from what happens when we write or speak. What we do not do with language is “dwell in already elaborated signs and in an already speaking world” and simply “reorganize our significations according to the indications of the signs.” In this rendering, there is no space for probability, because all possibilities have already been determined. In Bourdieu’s distinction, this is more accurately described as an apparatus that requires no motivation on the part of actors in order to use it, because objective rationalization so completely displaces all uncertainty.
This is a false picture, however, because a structure tends to ignore certain minor facts like what happens when we speak, and that as we do we engage not with pre-established signs the meaning of which is removed from probability. As Merleau-Ponty puts it, “to speak is not to put a word under each thought; if it were, nothing would ever be said.” Instead, language constitutes a space of objective probabilities that our action (speaking, writing) can be oriented to and which, in this case, acquires its meaning at least partially through the words we use to evoke our thought and make it actual (just as Matisse uses art to make his painting actual). Thus, the meaning of what we speak and write is at least partially determined by how probable or improbable our words are, what else we could have said given the objective probabilities of language at a given historical time, plus an interlocutor who has likewise learned those probabilities.
Both language and art are objectively constituted by probability, and when we engage with them, we therefore “take chances” and engage with uncertainty. That uncertainty can widely vary both subjectively and objectively. When we have not (yet) formed expectations, or learned objective probabilities, we might experience lots of (subjective) uncertainty, even if what we are doing is not improbable (objectively speaking). By the same token, an objective uncertainty can apply to what we do because what we do is improbable (at least at the current moment). If we return to non-compositional art, we can better appreciate what it means to invite chance in as, in this case, a technique to interrupt a kind of self-feeding loop that occurs when expectations so perfectly match probabilities that it can only result in, as Bois puts it, “the communication of form.” To engage in non-compositional technique, then, breaks habitus by making chance decide what will happen next, where the hand will go next, what will be the next result, rather than a finely-attuned set of expectations that can, essentially, do nothing unexpected in relation to what is objectively probable.
Bois, Yve-Alain. (1992). “Ellsworth Kelly in France: Anti- Composition in its many Guises” in Ellsworth Kelly : the Years in France, 1948-1954. National Gallery of Art.
_____. (1999). “Kelly’s Trouvailles: Findings in France” in Ellsworth Kelly: The Early Drawings, 1948-1955. Harvard Art Museum.
Bourdieu, Pierre. (1988). Homo Academicus. Stanford UP.
_____. (2000). Pascalian Meditations. Stanford UP.
Deleuze, Gilles. (2003). Francis Bacon: The Logic of Sensation. Continuum.
Dreyfus, Stuart. (2004). “The Five-Stage Model of Adult Skill Acquisition.” Bulletin of Science, Technology & Society 24: 177-181.
King, Antony. (2000). “Thinking with Bourdieu Against Bourdieu: A ‘Practical’ Critique of the Habitus.” Sociological Theory 18: 417-433.
Krauss, Rosalind. (1977). “Notes on the Index: Seventies Art in America. Part 2.” October 4: 58-67.
Merleau-Ponty, Maurice. (1964). “Indirect Language and the Voices of Silence” in Signs. Northwestern UP.
Singerman, Howard. (2003). “Non-Compositional Effects, or the Process of Painting in 1970.” Oxford Art Journal 26: 125-250.